BeBeryllium Research Symposium: Basic Mechanisms and Human Health
June 25- 26, 2002
National Library of Medicine, Bethesda, MD





 Session 1

 Session 2

 Session 3

 Session 4

 Session 5


Sponsored by
Department of Energy Seal
Office of Biological and Environmental Research,
The Department of Energy
in cooperation with


The National Institute for Occupational Safety and Health, and

National Jewish Medical and Research Center
The National Jewish Medical and Research Center


Session 4. Gene–Exposure Interactions

Functional TNF-a Promoter Polymorphisms Associated With Chronic Beryllium Disease and Beryllium-Induced TNF-a

L.A. Maier1, R.T. Sawyer1, T. Hendry-Hofer1, C. Parsons1, R. Bauer1, D. McGrath2, P. Lympany2, R. duBois2, and L.S. Newman1
1National Jewish Medical and Research Center, Denver, CO
2Imperial College of Science and Technology, London, UK

Background: Beryllium (Be) stimulates high levels of TNF-a from chronic beryllium disease (CBD) bronchoalveolar lavage (BAL) cells. Previously we have shown that Be-stimulated TNF-a production is associated with a G to A transition in the TNF-a promoter located at position -308 and with markers of disease severity. Although not statistically significant, lower Be-stimulated TNF-a was produced from subjects with a C to T transition at -856 in the TNF-a promoter. We questioned whether the functional -308 polymorphism was a risk factor for CBD and whether other TNF-a promoter variants were associated with high Be-stimulated TNF-a production.

Objective: This study tested the hypothesis that the -308 A TNF-a promoter polymorphism was present at a higher frequency in CBD compared to beryllium-exposed non-diseased subjects (Be-non-diseased).

Methods: Following informed consent, demographic information was obtained from CBD (n=85) and Be-non-diseased subjects (n=67) subjects. Genomic DNA was extracted from peripheral blood cells. A Taqman allelic discrimination assay specific for the -308 variants was used to determine the subjects’ genotypes. Direct sequencing of the TNF-a promoter from +64 to -1045 was used to confirm the Taqman results and detect other known TNF-a promoter variants, including those at -238, -376, -572, -856, -862 and -1031, in CBD subjects (n=43). HLA-DPB1 and -DRB1 genotyping was determined by sequence specific PCR in a subset of the cases (n=31). Bronchoalveolar lavage cells were obtained from CBD subjects (n=31) and cultured in the presence and absence of beryllium sulfate. TNF-a supernatant concentrations were measured from the BAL cells by ELISA.

Results: The -308A variant was found at higher frequency in CBD subjects (34%) compared to the Be-non-diseased subjects (19%, p=.04), with an odds ratio of 2.2 (95% confidence interval 1.01, 4.57). High TNF-a production was associated with the -308 A variant (median 6,000 pg/ml vs. 1,142 pg/ml from -308 GG variants, p=0.01) and the -856 CC variant (median 2,636 pg/ml vs. 528 pg/ml from -856 T variants, p=0.04), but not with the -238, -376, -572, -862 or -1031 TNF-a promoter variants. Individuals homozygous for HLA-DPB1 with a glutamic acid at position 69 (Glu69) produced higher beryllium stimulated TNF-a (median 6113 pg/ml vs. 1320 pg/ml from nonGlu69 heterozygotes or homozygotes, p=0.04). HLA-DRB1 DR2, DR3 and DR4 genotypes were not associated with Be-stimulated TNF-a production in CBD.

Conclusion: The -308 A variant (AA or AG) is a risk factor for disease susceptibility in CBD. It is likely that this association is due in part to the functional role the -308 polymorphisms plays in TNF-a production. Whether the association between the -308A variant and CBD is related to an extended TNF-a haplotype, including the -856 CC variant, and/or association with the Glu69 genotype, will require further investigation.

K08 HL03887, R01 ES06358-06, M01 RR00051


Association of HLA Class II Markers and Beryllium Hypersensitivity and Disease: Reevaluation of Previously Published Data after 4-8 Years of Follow-up

C. Saltini and M. Rossman
Malattie dell'Apparato Respiratorio, Università di Roma “Tor Vergata”, Divisione Clinicizzata di Malattie Respiratorie, Istituto Nazionale per le Malattie Infettive, “L. Spallanzani”-I.R.C.C.S. via Portuense 292, Roma 00149

After the first report of the association of the genetic marker HLA-DPB1 Glu-69 with susceptibility to chronic beryllium disease (CBD), a number of observations have associated beryllium hypersensitivity with a number of HLA locus genetic markers. In particular, HLA-DPB1*0201, *0601, *0901, *1001, 1601, 1701, 1901 and HLA-DQB1 Gly86 have been associated with CBD, HLA-DRB3 Arg74 and HLA-DPB1 Glu-69 with susceptibility to beryllium sensitization, and TNF-a -308A with beryllium hypersensitivity in the presence or the absence of granulomatous disease. However, the studies published so far are limited by: 1) the small size of the population groups, sometimes too small to assess associations with less frequent HLA-DRB1, B3, B4, B5, HLA-DP rare alleles; 2) a follow-up period insufficiently long to assess progression from sensitization to disease with a condition known for the long latency from exposure to clinical presentation; 3) non uniform phenotypic description i.e., lack of uniformity in the use of descriptors such as  “lymphocytic alveolitis,” “positive BAL Be-LPT,” or lacking pulmonary function data as diagnostic criteria. These limitations hamper the ability of investigators to combine data from different studies to make a comprehensive assessment of the strength of the association of genetic markers with sensitization and/or disease. The aim of the present study was to assess the association between HLA class II (HLA-DP, DQ, and DR) genetic markers and beryllium hypersensitivity. To determine if the association is between these markers and beryllium hypersensitivity, beryllium hypersensitivity without disease or with clinical chronic beryllium disease. Subjects with beryllium hypersensitivity and exposed controls that have been previously published will be reevaluated. Follow-up studies will be performed to determine whether these subjects progressed to CBD or had beryllium hypersensitivity without clinical disease. The resulting population that was comprised of 62 hypersensitive subjects, 53 diseased subjects and 175 control subjects. Both populations, which had been HLA class II typed with standard molecular methods, were re-characterized using a common phenotypic criteria. Follow up will be from 4-8 years.


Electrostatic Potential on HLA-DP*1701 and HLA-DP*0401: Implications for Putative Mechanism of Chronic Beryllium Disease

James A. Snyder, Ainsley Weston, Sally T. Tinkle, and Eugene Demchuk
Health Effects Laboratory Division, National Institute for Occupational Safety and Health, Morgantown, WV 26505-2888, USA    

Presentation in html or PowerPoint

The pathobiology of chronic beryllium disease (CDB) involves MHC class II human leukocyte antigen (HLA). Molecular epidemiological studies indicate interaction of specific HLA-DPB1 alleles as a factor in disease susceptibility. We have studied three-dimensional structural models of HLA-DP proteins encoded by these genes. The extracellular domains of HLA-DPA1*01031/B1*1701 and HLA-DPA1*01031/B1*0401 were modeled from the X-ray coordinates of HLA-DR as a template. Using these models the electrostatic potential at the molecular surface of HLA-DP was calculated and compared for both isoforms. These data show the distinguishing characteristics of the isoforms in the vicinity of the antigen-binding groove. The presence of a positively charged lysine in position 69 in HLA-DPB1*0401 was compared with the negatively charged glutamate in the same position in the 3D-structure of HLA-DPB1*1701. We calculated the pKa values for the titratable sites in these proteins. Using these predicted values we estimated the mean charge on relevant titratable residues at physiological pH, and assigned an approximate total charge to the protein. The total charges based on the standard pKas for amino acid residues are -17/-9 and -17/-3 for HLA-DPA1*01031/B1*1701 and HLA-DPA1*01031/B1*0401, respectively (and compared to -14/-6 on the HLA-DR template). Hence, the combination of CBD-susceptible alleles HLA-DPA1*01031/B1*1701 may encode proteins that carry the largest negative charge. The protein charges based on the calculated pKa values are lower, but the highest negative charge still remains on HLA-DPA1*01031/B1*1701.

The calculation of pKa values also permits examination of the interaction energies between the titratable sites. Specific pairs of residues in proximity to the binding pocket are shown to have large interaction energies, indicating electrostatic coupling between these residues. The majority of these residues are on the B chain encoded by the HLA-DPB1*1701 and HLA-DPB1*0401 alleles, and 6/7th of the difference in the total negative charge results from substitutions in positions B55, B56, B69, B84 and B85. Interestingly, the results of epidemiological studies have implicated these positions as being potentially important for conferring susceptibility to either CBD [1-3] or to other hard metal lung disease [4]. These results are useful for interpreting the molecular recognition in MHC class II proteins, and for investigating its interactions with Be2+ in chronic beryllium disease.

1. Wang Z., et al. (1999) J. Immunol. 163, 1647-53; 2. Wang Z., et al. (2001) Toxicology 165, 27-38; 3. Rossman M.D., et al. (2002) Am. J. Respir. Crit. Care Med. 165, 788-94; 4. Potolicchio I., et al. (1999) Eur. J. Immunol. 29, 2140-7.


Original Announcements: Agenda - Symposium Information - Speakers - Symposia - Hotel Accommodations - Registration Form - Abstract Submission Form - Poster (pdf, 1365 KB)